diode v-i curves

millwood

with some free time today, i decided to try to plot out the v-i curves for some of the more common diodes. the part I am interested is am detector so I am more focused on small signal levels.

I tried a few rectifying diodes (1n400x), some leds and then some signal diodes / transistors.

here is the part of signal diodes (in this case, 1n4148, and 2n5551's b-e junctions).



the two curves on the bottom are for two 1n4148s from ONSemi. as you can see, they track each others very closely.

the blue and the green lines are for a 2n5551's b-e junction (the blue trace) and then as an amplified diode (bc-e junction, the green trace). As you can see, the v-i curve is much flatter with b/c tied together.

millwood

here are v-i curves for some leds.



both red and bright yellow leds are pretty good as a voltage reference. QDSP-3892 is a HP red 7-segment led.

millwood

similarly, here are the v-i curve for the rectifying diodes (and a couple zeners).

1n4006/4007 are highly alike. and the zeners (1n5929/1n5927) are highly alike.

millwood

btw, the Y-axis is mv, and X-axis is ma.

millwood

I have a few schottky diodes and fancier diodes (sd101, bat54 and hsms). once i have them mounted, I will plot them out as well.

millwood

the behavior of 1n4148 is a little bit surprising to me in that it opens like 250mv or so (@ If = 1ua), making it almost like a schottky diode. I guess schottkies can potentially start conducting at even lower Vfwd at a comparable If.

millwood

when you simulate a circuit, you often wonder how accurate those models.

here, I used ltspice's 1n4148 model and plotted its theoretical v-i curve (in green) vs. the measured 1n4148 v-i curves (two thinner lines as I had measured two 1n4148s).



they are right on top of each other. the error rate is about 2%.

Wow! you have to give ltspice credit for being this accurate.

葛优

常见的二极管起始角曲线图不错值得收藏。

yngz

问一下，加有正向偏压的检波二极管，虽然减小了其起始导通电压，但反向漏电也等效地增大了吧。是不是会减小检波电路所在LC回路的Q值，反而降低其检波效率？
所以，使用起始导通电压低，反向电阻大的二极管检波效率最高。

老国哥

都怪当年没认真读书

millwood

问一下，加有正向偏压的检波二极管，虽然减小了其起始导通电压，但反向漏电也等效地增大了吧。

yes.

是不是会减小检波电路所在LC回路的Q值，反而降低其检波效率？

yes. However the reverse leakage is so small that I doubt it will have any material impact.

所以，使用起始导通电压低，反向电阻大的二极管检波效率最高。

generally true. However, sky high Q isn't that good. What you want is a diode with Rd in the 50k - 150k range - most germanium diodes are in that range, and most silicon / schottky diodes are in the 1Meg or more range. LEDs tend to be 1x10^15 or higher.

I think germanium diodes' performance has far less to do with its Vfwd or Ir but more to do with its Rd parameter.

yngz

Rd值是二极管在零压差下的等效正向电阻，它与二极管在小信号检波下所产生的非线性失真有关吧。Rd值越大，与二极管达到起始导通电压时的导态电阻的差别就越大，即二极管在起始压差下的非线性导电效应就越显著，因而检波的非线性失真就越大，在这个区域产生所谓的平方律检波。对于理想二极管，Rd值为零，不管信号强弱，都可以实现完全的线性检波，效果最好。当然实际的二极管远非理想二极管，适宜的Rd值对于小信号可以实现最佳的平方律检波。Rd值太小或太大，平方律检波的效率都会变差。

e3po

As far as I could recall,  you mentioned that you got 1000pcs of 1n5711.

Why not plot one of them as well ?


I'm very interested in the 1n5711, coz I've never got my hands on any.

millwood

haven't had time yet. maybe this week. stay tuned.

老勐

老国哥 发表于 2012-9-23 01:53
都怪当年没认真读书

楼主很热心地给咱们测试了一些半导体PN结的正向伏安特性。有常见的1N4148，也有三极管的PN结，还有几种发光管的正向特性。谢谢他。